Identifying candidate genes; theory and practice

So, we, along with collaborators here at Penn State, and other institutions, happen to be working on a project to map genes associated with craniofacial morphology in mice and baboons.  Yes, you can laugh, given all we've said about the problems with looking for genes for traits, that we too are involved in genomewide mapping!  But, genes are involved in morphology, and mice and baboons do differ, perhaps even systematically (by lineage), so, in theory, it's not a futile search.  And, we're hoping that all our skepticism about how these kinds of studies are usually done can point us toward more useful ways to think about the problem--not to enumerate genes but to document or better understand complex genetic causal architecture.  In theory.   

We're at the point now where traits have been measured on about 1200 mice, and these mice have been genotyped at thousands of variable nucleotide marker sites across the genome, the statistics have been run, 'lod' scores generated, and we're ready to identify candidate genes associated with these traits.

The lod scores -- measures of the probability that a genetic marker, a variable locus along a chromosome, is near to a gene that contributes to variation in a trait -- can point to fairly wide or fairly narrow intervals on a chromosome.  The scores of interest are those that are statistically significant, or 'suggestive', that is, almost statistically significant.  Some intervals indicated by statistically interesting marker sites are in gene-rich regions, some in gene-poor regions of their respective chromosomes, so some are easier to mine than others.

What we do with each chromosomal interval is to identify all the genes along it, and then hunt for data on how each gene has been characterized.  We are doing this manually, but it can be, and often is automated, with a program that looks for certain kinds of genes, in specified gene families, for example.  We are choosing to do this by hand because we are trying to reduce the preconceived assumptions with which we approach the problem.  Our goal is to collect expression and function data about each gene in every interval.  Yes, that's a lot of genes, but we know that most genes are pleiotropic, expressed in many tissues, with multiple functions, most genes have not been fully characterized, and we don't want to assume things a priori, such as that a given class of gene is what we're looking for. 

But this is quite problematic. 

Let's take an interval we slogged through yesterday.  It's one that mapping of several different traits yielded significant lod scores for, so potentially of greater interest than the intervals that were only tagged by one trait.  It may, for example, reflect multiple traits that have correlated genetic causation during the development of the head.  Anyway, this interval, on mouse chromosome 8, includes 95 identified genes, and a few identified micro RNAs (DNA coding for short RNA molecules that affect the expression of other genes).  In fact, and not unusually, very little of the non-coding DNA in this region has been characterized, which is problematic itself, since some of that is surely involved in gene regulation, and who's to say that the variation we're looking at is not due to gene regulation.

Of these 95 genes, we can't find expression data for 49 of them.  Over half.  So, are they expressed in the anatomical region relevant to the head traits that mapped to this area?  We don't know.  It's a major task to study the expression of even one gene in mice, much less a host of genes.  Apparently no functional data are available for 15 of these genes, and of the rest for which function has been characterized, many are assumed to have a given function just because its DNA sequence is similar to a gene whose function is known.  Many others are characterized as "may be involved in" such and such. And, a handful have been identified to be involved in a specific function, but expression data show that they are expressed in other tissues or regions of the body -- this seems to be particularly true for 'spermatogenesis' genes, which tend to show up everywhere.*  Olfactory (odor) receptors as well.

So, given the inconsistencies, vagaries, and incompleteness of the data, how are we supposed to go about choosing candidate genes?  One way is the Drunk Under the Lamp-post approach (looking for lost keys where there's light even if that's not where they were dropped):  to grab a gene we already know about, or narrow our search to genes in families that make some kind of biological sense. That's looking for things we already know and making the comforting assumption they must be the right ones.  This will clearly work sometimes, and may be better than totally random guessing.  But, such an approach of convenience perforce ignores many plausible candidates, both that have been characterized, and those that have not.  So, even when genomic databases are complete (if that day ever arrives -- but keep in mind that different patterns of gene expression are associated with different environments, and thus that characterization may never be complete; reminding us of Wolpert's caution about yet another knock-out mouse with no apparent phenotype; "But have you taken it to the opera?") this approach may not identify all actual candidates.

And this is all true even if we believe we're looking for single genes 'for' our traits.  What about when we throw in the kinds of complexity we are always going on about here on MT?  It's fair to say that most gene mapping studies don't result in confirmed genes for traits, or a few are found but usually with relatively minor contribution relative to that of other, unidentified genes.  See most GWAS, e.g.  And if we had access to unpublished GWAS results, this would be even more convincing.

So, while we do believe that genes underlie traits in some collective aggregate ways, and that in theory they are identifiable, current methods, and the current state of (in)completeness of the databases leave us more convinced than ever that identifying genes 'for' traits is an uphill slog.

-----------------------------------------------------------------------------------------------------------------------

 

* Here's a section from an in situ experiment on a mouse embryo, taken from the extensive gene expression database, GenePaint.  The darker areas are where the gene in question is expressed.  In this case, the strongest expression can be seen in the brain, the pituitary gland, the lungs, the whiskers, the growing sections of the digits, and there is expression in other tissues as well.  The function of this gene is characterized as, "Probably plays an important role in spermatogenesis." 

How complex is 'complex'?

The word 'complex' is frequently used, though not always as clearly as it might be. In today's genetics arena it means a trait that is the result of multiple genetic elements as well as environmental factors that are usually unknown or not specified, but can include the genetic element's genomic background. Can we get a clearer understanding of this interaction in some way that has not yet been well-explored?

Most complex traits, whose genetic contributors GWAS and related mapping methods are designed to find (see earlier posts) show substantial evidence of being 'genetic' in some sense: there is correlation of the trait among relatives or an association of risk of the trait--like a disease--among family members.

The problem is that despite evidence for genetic involvement, GWAS and other methods have only been able to identify a small fraction of the contributing elements. One response is that we need larger studies. Another is that the objective is not to account for the disease in terms of genes, but to find genetic pathways that are involved.

Most common diseases have increased substantially, if not dramatically, within living memory and more importantly within the time since trustworthy epidemiological data on incidence (rate of new cases per year) or prevalence (fraction of persons affected) have been available.

This would suggest to reasonable people, even including some geneticists, that at least for preventive purposes the major responsible (and avoidable) factors for the disease are environmental, such as exposures to risk factors like toxins, lifestyle changes such as in diet, etc.

A few years ago, the molecular technology infrastructure for mapping studies was laid down, and paid for on the rationale that common genetic variants were likely responsible for these common diseases--and hence that genetics was a right way to approach them. Common variants for common disease (CVCD) became a mantra.

In response to the environmental and other arguments raised even at the time, proponents of CVCD and the investment in the gene-mapping infrastructure (e.g., the HapMap project) said that, yes, environmental factors clearly were involved, but the increase in prevalence was due to their interaction with common genetic susceptibility variants.

Subsequent mapping, including numerous, often huge genomewide association studies, has generally failed to find such variants. The meaning of 'common' can of course be adjusted to fit results, but the bulk of the heritability of these many studied traits remains unexplained. It's a fair question whether these traits are truly complex and largely unmappable, or whether we just haven't studied them enough.

A kind of widespread relevant evidence may be the following. The substantial heritability of common disease as well as normal traits suggests that many genes contribute; the traits are often called 'polygenic' for that reason. But these many genes might individually vary in their effects. For many theoretical and empirical reasons, one would expect some alleles (genetic variants) at one or a few genes, to interact or respond more strongly to changing environmental factors.

If that is the case, then the more important genes that were not identifiable in case-control or family samples before the environmental change, should be mappable afterward. That's because those variants that would be the main responders to the environmental change, whatever it was. Their individual effects, modest before the change, should be major after it.

Yet, today, after a long list of diseases have had large, rapid increases in prevalence, the GWAS findings are as we have seen: they are not identifying much that is of population-scale importance. On the surface, this suggests that the argument about complexity really is correct: there are, indeed, many genes involved, but they each make very small net effect on risk. A few are detected whose effects are greater, but they are few and even their effects are only modest.

From this perspective, which is based on data, not theory, secular trends in risk and the failure of GWAS to find CVCD's is relevant data, suggest that complex traits really are basically homogeneous in terms of genetic causation.

Now, if this is true it constitutes material evidence that should change our understanding of the nature of these traits: why would it be that there are generally no major alleles waiting for environmental changes to give them a chance to be expressed? Indeed, isn't that just how natural selection is supposed to work, with environmental change favoring 'good' genetic variants in the population and raising them to high frequency? Those variants should have substantial effect on the trait so the organism carrying the variants would reproduce more successfully.

If our thinking is correct, then this tells us something. Perhaps the networks of which biological traits are built are internally adjusting--strong changes in one part of a pathway network lead to slowing down of others. Yet, secular trends show that the net result can involve major change. It is indeed somewhat difficult to believe that the genetic responses to environmental changes are so internally homogeneous that even after major stimulus none really stands out even when studied in large samples. There must be a message there--if we can but figure it out!

These are just superficial ideas at this point, but they could help direct changes in what we look for, or how we look. We are starting to use an evolutionary simulation program that Brian Lambert in our group has written (see the description of ForSim on Ken's web page for details) to see if this point is correct as we think, or if there is some aspect of genetic control that we are overlooking. Stay tuned for results.